Scientific Highlights of AIDS Vaccine 2013, Barcelona
AIDS Vaccine 2013 was hosted by the Global HIV Vaccine Enterprise together with the Catalan Program for HIV Vaccine Research (HIVACAT) at the Barcelona International Convention Center from 7 – 10 October. More than 900 researchers, funders, advocates and policymakers met in Barcelona for the 13th AIDS Vaccine conference at a time of significant progress and challenges in HIV prevention research.
The conference opened and closed with discussions around HVTN 505, a phase 2b clinical trial stopped earlier this year when the vaccine tested was found to be ineffective. Researchers discussed the perplexing results from this as well as other emerging clinical trial data and new analyses of existing trial data to continue to better predict complex vaccine-induced immune responses that can help ensure protection and teach us how to design our immunogens to optimize these responses. The AIDS Vaccine 2013 conference showed us once again that our setbacks will continue to inform our successes.
All sessions are available via audio and webcast at http://www.aidsvaxwebcasts.org/. Conference abstracts were published in an open access issue of AIDS Research and Human Retroviruses and published on its web site: http://online.liebertpub.com/doi/pdf/10.1089/aid.2013.1500. An electronic, user-friendly, open access archive of posters (e-posters), hosting close to 50 percent of presented posters, is available by visiting the web site: http://epostersonline.com/aidsvax2013/.
The conference theme of Progress, Partnership and Perseverance was at the core of the program with the presentation of more than 500 abstracts containing new data informed by decades of scientific progress, new partnerships and consummate dedication and perseverance. Highlights included more details about the structure of the HIV envelope, the pathway for the formation of broadly neutralizing antibodies and a number of innovative approaches to determine just how to induce these neutralizing antibodies. Novel vaccination strategies being investigated in non-human primates, including a new T-cell immunologic paradigm through studies with the CMV vector and other viral and non-viral vector vaccine strategies provided much hope and presentations on innate immunity, including the role of adjuvants, and the use of systems biology approaches to inform rationale vaccine design, added to the optimism at the meeting. This report highlights some of the noteworthy updates shared during the conference.
Dr. Magdalena Sobieszczyk, Columbia University, provided an update on the HVTN 505 trial, which tested the efficacy of a DNA prime-recombinant adenovirus type 5 boost (DNA/rAd5) vaccine regimen at the opening session “Insights from HIV Vaccine Efficacy Trials: What We Are Learning from HVTN 505.” The trial was stopped in April due to futility of both of the primary end points: prevention of infection and reduction in viral load. The interim analysis of viral loads in breakthrough infections showed no statistical differences between the placebo and the vaccine arms. The analysis of infections occurring after week 28 post-enrollment showed 27 infections in the vaccine arm and 21 in the placebo arm. Modified-intention-to-treat (MITT) analysis, which includes all infections occurring during the study, showed 41 infections in the vaccine arm and 31 infections in the placebo arm. While the differences were not statistically-significant in both analyses, they raised the question of the possible vaccine-associated enhancement of HIV infection. Post-unblinding follow up of trial participants showed the cumulative number of infections in the MITT cohort was 46 for the placebo arm and 40 for the vaccine arm. Thus, no enhancement of infection was observed in the extended study data. Ongoing work focuses on understanding the immune responses elicited by the vaccine.
Dr. Genevieve Fouda, Duke University Medical Center presented “HIV-1 gp120 Vaccination Elicits a Robust and Durable Anti-V1/V2 IgG Response and Yet No HIV-1 Env-Specific IgA Response in HIV-Exposed Infants.” She measured antibody responses elicited in two previously completed phase I pediatric HIV vaccine trials (PACTG 230/ PACTG 326) showing that infants were eliciting cross clade Env-specific IgG and anti-V1V2 IgG (IgG3) responses. At week 52 93% of infants vaccinated in PACTG 230 still had anti-V1/V2 IgG responses, as compared to 64% in RV144 vaccinees at peak immunogenicity. These data suggests that HIV vaccination in infants may be as potent as adult vaccination in eliciting strong antibody responses.
Dr. Glenda Gray, Perinatal HIV Research Unit in South Africa, gave “An Update on the Phambili/HVTN 503 Study, a Phase IIB HIV Vaccine Efficacy Study Investigating MRK Ad5 HIV-1 Subtype B gag/pol/nef Vaccine.” Immunizations were stopped in the Step and Phambili studies in 2007 after the data and safety monitoring board (DSMB) for Step found that the vaccine did not provide protection. An analysis of the Step data at the time the trial was stopped in 2007 found that certain subgroups enrolled in that trial (men who had previously been exposed to Ad5 and uncircumcised men) may have had an increased risk of HIV infection, particularly in the first 18 months following vaccination. No trend towards higher infection risk was initially seen in the Phambili participants; however, Dr. Gray presented data showing an increase in infections among vaccinees in the longer-term follow-up. This effect was seen more in men than in women and may persist for longer that was observed in Step. More studies are planned to understand the mechanism behind the increased infections.
Dr. Peter Gilbert, SCHARP, presented a meta-analysis combining data from the three efficacy trials using the adenovirus 5 vector, Step and Phambili along with the HVTN 505 trial, at his presentation entitled “Meta-Analysis of Ad5-vector HIV Vaccine Trials to Assess the Vaccine Effect on HIV Acquisition.” This analysis showed that the HIV incidence in placebo recipients was 2.1% and among the vaccine recipients 2.9%, resulting in a 33% greater risk of HIV infection among vaccine recipients compared to placebo recipients. Unsurprisingly, when data from the HVTN 505 trial was excluded (no trend towards higher infection in vaccinees was observed in HVTN 505) from the analysis, the HIV incidence was 2.2% in placebo and 3.1% in the vaccine arms, increasing the risk of infection of vaccinees to 41%.
Three groups reported on their work on different herpes viruses to create live-attenuated, persistent vaccines that are currently being tested in NHPs. Despite the differences, all three vaccines elicit effector memory responses and some vaccinated animals show similar phenotype across the studies with an apparent systemic SIV infection that is later completely controlled (and in some cases, eradicated) by the immune system.
Dr. Louis Picker, Oregon Health & Science University, provided an update on the NHP studies of a vaccine based on replication-competent cytomegalovirus (CMV), which expresses SIV antigens – “Pre-Clinical Development of Cytomegalovirus Vectors.” He previously reported that vaccinated macaques divide into two groups upon vaginal challenge, 50% of animals being just as vulnerable to SIV as animals receiving placebo and the other 50% showing full protection. Protection is observed after a small temporal “blip” in viral load in blood and peripheral tissues, indicating productive infection, which later gets completely controlled and, moreover, cleared. In the new work presented, these findings were extended by challenging the vaccinated macaques with intravenous injection of a large amount of SIVmac239. Again, a similar protection phenotype was observed in approximately half of vaccinated animals. Extensive search in the protected animals for the virus 1.5–3 years after infection showed very small amounts of detectable viral genomes, and no replication-competent virus could be recovered from these animals. Currently Picker’s group is exploring the applicability of such a vaccine to TB as well as studying the very unusual T-cell responses elicited by the CMV-based vectors, which include MHC class II restricted CD8 cells with very broad specificity. They are also testing these CMV-based vaccines in therapeutic settings in NHPs and are working on creating a humanized version of the CMV vector, focusing on retaining immunogenicity while ensuring safety.
Dr. Ronald Desrosiers, New England Primate Research Center, Harvard Medical School, presented data on a vaccine based on rhesus monkey rhadinovirus (RRV), a simian herpesvirus establishing a persistent infection in B cells — “Impressive Vaccine Protection Against SIVmac239 by Recombinant Gamma-2 Herpesvirus.” An RRV-based vector was designed to express SIV tat/rev/nef, gag and env. Previous experiments showed that similar to the CMV-based vector used by Louis Picker’s group (both vectors are persistent herpesviruses), RRV also elicits effector-memory T cells, but the responses are targeted to the immunodominant epitopes in SIV. Vaccinated animals showed 1.7-2 log reduction in viral load as compared to animals receiving placebo. However, no antibodies to Env were detected. Their group addressed this issue by optimizing Env expression by codon-optimization of the env gene. Animals vaccinated with the redesigned vaccine developed the antibody responses to Env. When repeatedly challenged with SIVmac239, three of six animals showed “blip and complete control” phenotype, very reminiscent of the protection observed by the Picker group in their CMV trials. Two other animals got infected followed by 2–3 logs reduction in viral load and one of the vaccinated animals remained uninfected.
Dr. Kelly MacDonald, Mt. Sinai Hospital, University of Toronto, described the work with varicella zoster virus (VZV), a human herpesvirus that establishes a persistent infection in neural ganglia — “Protection of Cynomolgus Macaques from pathogenic SIV following vaccination with Varicella-Zoster virus based vaccines.” Her group used human VZV as a replicating vector carrying SIV antigens to vaccinate macaques, in which the virus is attenuated, against SIVmac239. Similar to the CMV- and RRV-based candidates, the vaccine elicited strong effector memory responses, although not in all of the animals. When these animals were challenged with SIVmac239, all eight were infected as evidenced by significant peak viral load, but three animals quickly controlled the virus to undetectable levels (one out of nine animals became a controller in the placebo group). Effector memory T cell responses correlated with the observed protection, suggesting that further work needs to be done to improve these responses. One potential advantage of the VZV-based vaccines is that an attenuated VZV vaccine has been licensed and used in a large number of people.
Dr. Ian Wilson, Scripps Research Institute, presented on “Structure of a soluble, cleaved HIV-1 Env trimer and its recognition by broadly neutralizing antibodies,” adding to the effort to unravel the structure of the envelope and its interaction with broadly neutralizing antibodies. He showed how they were able to improve the resolution of the structure using cryo-electron microscopy reconstruction and x-ray crystallography of a fourth generation cleaved, soluble SOSIP gp140 trimer (BG505 SOSIP.664) bound to PGT122 Fab. There was good agreement between their cryo-EM structure and x-ray structures. The HIV Env trimer, structure reveals the spatial arrangement of Env components, including the V1/V2 and V3 loops, HR1 and HR2 domains, and shielding glycans. The V1/V2 conformation agrees overall with scaffolded V1/V2 in complex with PG9/PG16, the structure of the V3 loop is also similar but the position is different as compared to x-ray and NMR structures. The structure also provides insights into the PGT 122 epitope and the glycans involved, which covers a more extensive area than previously described.
Finally, a number of presentations focused on immune responses and therapeutic modalities at mucosal and epithelial sites in light of evidence that HIV is most vulnerable to immune control immediately following the transmission event and before systemic dissemination of the infection.
Dr. Behazine Combadière, the University Pierre and Marie Curie in Paris, presented a proof of concept clinical trial eliciting different immune responses through hair follicle vaccination (Cutaneous Routes for HIV Vaccination Strategies). The hair follicle vaccination route is realized by injection of immunogen coated nano-particles that diffuse and penetrate through the hair ducts and distribute differentially according to their size. The vaccine is administered via three different skin routes (intra-dermal, sub-cutaneous or through hair follicle targeting) and each route leads to a different immune pathway. Dr. Combadière presented data were either a nanoparticle bound to HIV-gag or MVA-HIV-gag vaccine was injected at the three skin sites previously shown to activate different immune pathways. The intradermal vaccination route (which corresponds to follicular targeting) recruited the wider range of immune responses with a strong antigen-specific CD8 T cell response mediated by Langerhans cells and was further scrutinized. Upon intradermal injection of the nanoparticle vaccine, activated cells loaded with nanoparticles migrate from the epidermis to the dermis and into lymph nodes within hours. There they are responsible for the induction of a CD8 T cell response and partially involved in the induction of T follicular helper cells. Intra-dermal immunization with the MVA control or the MVA-HIV-P24gag vaccine induced the priming and proliferation of CD8 T cell in the bone marrow independently from the blood compartment. The authors showed that the priming of these CD8 T cells in the bone marrow is mediated by antigen presenting neutrophils that migrated from the draining lymph-nodes. Dr. Combadière argued that these studies demonstrate the feasibility of hair follicle vaccination and show that this type of injection leads to the induction of two independent types of effector CD8 T Cells respectively in the bone marrow and in the draining lymph-nodes. It is not known at this time if these two populations have distinct immune activities during the HIV infection or if they exert protection, but this result suggests that the combinations of several vaccines and routes of immunizations may be the best way to recruit the wider breadth of immune response pathways and this is being tested in the CUTHIVAC EU FP7 trial.
Dr. Ashley Haase, University of Minnesota, presented a study describing the events at the cervical mucosal barrier during high dose SIV challenge in the rhesus macaque model — “Targeting the Female Reproductive Tract Epithelium to Prevent Transmission and Acute Systemic Infection: Lessons Learned in the SIV-NHP Model.” Early events of transmission at sites of exposure are greatly scrutinized as potential targets for prevention tools. Dr. Haase argued that good NHP models for early infection are available and can be used to identify correlates of protection at mucosal sites and inform on vaccine design and evaluation. Similar to HIV infection in humans, the SIV founder population is quite small and the recruitment of target cells (mainly CD4+ T Cells) to the site of entry is a necessary step for viral expansion and progression of infection. Dr. Haase presented data using a delta-nef SIVmac239 live attenuated NHP vaccine capable of inducing robust local protection to high dose vaginal challenges in macaques. Cell imaging reveals a small population of SIV RNA positive cells in the endo-cervix tissue 4 days post SIV challenge in non-vaccinated animals, followed by recruitment of CD4-T cells and rapid local viral expansion. In contrast, protected animals show less RNA positive cells at day 4, no CD4 T cell recruitment and no expansion at day 10. Protection is mediated by anti-gp41 IgG antibodies produced by vaccine-elicited local plasma cells. The anti-gp41 antibodies concentrate at the mucosal border where they bind neonatal Fc Receptors expressed on the cervical reserve cells. Anti-SIV antibodies at the epithelial border capture viral particles, block transcytosis, and help recruit anti-inflammatory pathways of the innate immune system and inhibit CD4 T cell expansion. This mechanism is the principal spatial and temporal correlate of protection in the cervical epithelium in this model.
Dr. Rupert Kaul, University of Toronto, presented prospective studies of innate and adaptive mucosal immune responses in highly-exposed sero-negative (HESN) individuals with a focus on defining immune correlates of both risk and reduced susceptibility to HIV infection at sites of sexual exposure — “Genital Immunology and HIV Susceptibility.” The immune response in genital tract mucosa was examined in two populations of HESN individuals: female sex workers (FSW) and men with long term HIV infected female partner not on ART. Cross sectional studies of immune responses in the genital tract of FSW showed elevated concentration of soluble antimicrobial factors such as elafin, serpins and defensins, as well as a decrease in pro-inflammatory signals. Other observations include: a higher concentration of IgA with neutralizing HIV activity in the genital mucosa than in the blood compartment, T cells immune quiescence in blood and female genital tract, a low frequency of HIV specific CD8 T cells and a decrease of Th17 T cells. Prospective analysis in a blinded study of PBMC and mucosal samples from HESN women showed that HIV infection was negatively correlated with anti-HIV IgA levels. In HESN men, studies on foreskin swabs and tissues collected post-circumcision revealed an increased level of anti-HIV IgA and of defensins in the foreskin; T cell immune quiescence was also observed (reduced IL-17 production, reduced TNFa) in the blood. A prospective longitudinal study of foreskin swabs from uncircumcised men showed the same association of protection. Dr. Kaul concluded noting that mucosal determinants of either risk or protection at sites of sexual exposure (vaginal, penal and rectal mucosa) are still poorly understood. Ex vivo models to test the projected correlates for either risk or protection are needed.